Loading

MakerBot Print is our newest print-prepration software, which supports native CAD files and STL assemblies,
allows you to interact with all your printers via the Cloud, and many other exciting new features.

Download Now

cfeniak

C-Bot 3D Printer

by cfeniak Oct 14, 2014
Download All Files

Thing Apps Enabled

Please Login to Comment

Hi, I nearly finished the build and its been a real joy and I am really happy with the way its gone...1 question I wanted to ask is in the firmware (Marlin) you are using Repatier Host and I want to use the RepRap discount smart controller and the last time I tried to install a Marlin file after adding U8glib and uncommenting the controller I got an error ... not tried this Marlin file but have you used with an LCD ...really grateful for any feedback ..Thanks John

Were you able to compile and flash your ramps successfully without the LCD enabled & U8glib library added?

Hi, Yes the program in your files compiled and uploaded succesfully, but Smart ramps LCD display lights up but no characters, I know I can add the U8Glib and uncomment the Smart ramps LCD but will that work, not tried any further supply to board got to check all wiring is correct and PSU works first..But Many Thanks John

Any chance you have the two connectors swapped on the back of the LCD?

Hi, connections are correct I have an Anet A8 I put Ramps 1.4 board and 12864 display without problems, but with this file if I uncomment Rerap smart discount I get this error... * \struct fpos_t

  • \brief internal type for istream
  • do not use in user apps
    */
    struct fpos_t {
    /* stream position /
    uint32_t position;
    /* cluster for position /
    uint32_t cluster;
    fpos_t() : position(0), cluster(0) {}
    };

I got a couple of other issues, one of the motor windings was wrong the RH motor was connected differently to skt so sorted that but know F2 on Ramps 1.6 board blown, not sure why no shorts and also find Extruder and Fan dont put volts on connector Heatmat is fine, I am using Mosfet on nozzle and SSR on Bed, also going to try another Marlin file...

My issue with no 12 volts was F2 blown because I had short on a connector to Nozzle mosfet, gonna put another board when it comes then still got issue with display ...

Hello,
Sorry for asking so many questions lately....
So if I want to make the version with triple lead screws for the Z-axis, is there any physical changes for the frame, or is it just a matter of printing some different parts? Also I want to drive all three lead screws with a single motor driven by a belt. I plan to use a NEMA-23 here so is there any modifications that I need to make here?
Thanks,
Taylor

Mostly the same parts. You need to add a cross member to the front of heated bed platform and perhaps one at the bottom of the frame as well to hold the drive motor. Otherwise you use the same parts and remixed prints for the 3 leadscrew setup. I haven't seen nema 23 version though. Would be easy to design a motor bracket for it though.

Hello,
Have you heard of anyone with a belt driven Z-Axis? I know there are disadvantages but has it been done?
Thanks,
Taylor

Belt driven leadscrews, but not straight up a belt elevated bed. I haven't seen it as a mod to this printer, but I have seen it on others. I don't see why you would though, leadscrews can be sourced for a reasonable price these days (not so 5 years ago).

Ah ok makes sense. It was just a thought that came to me for some reason.

Hello,
I am thinking about making a very large version of this printer. I am thinking about 550x550x650mm build volume. Because this is almost twice as large as the 12x12in that you stated in the description, would this work at this size? What changes do I need to make? Will it work with E3D Chimera bowden? And finally would you recommend this or the D-Bot?
Thanks,
Taylor Dettling

That is pretty substantial in size. Your challenges will be frame rigidity, squaring, and the Z axis. Use Aluminum plates where you can instead of the standard plastic printed ones. Use one of the many three leadscrew Z axis mods for the Z axis (see http://www.thingiverse.com/cfeniak/collections/c-bot). Keep the moving bar light (bowden). There is a remix Chimera carriage on Thingiverse, but I can't imagine not using a single or dual volcano hotend for this build volume.

The D-bot is a remix of the original C-bot, most parts are nearly identical and interchangeable so it doesn't matter. The biggest change was the Z axis and I don't recommend either the for your build size (triple recommended). I would recommend the D-bot endstops as they are nicer.
For the main upper corner brackets I would recommend the C-bot Remix as I beefed up the part thicknesses and you'll benefit from that.

All said and done, you may need to clamp down on print speeds due to vibrations. You printer size is ambitious to say the least.

When you say triple Z, should I do three lead screws driven by one motor with belts or three independent z-motors? I can see the Z axis getting out of sync but with a MIC-6 aluminum plate that’s ~16-18 in squared, I doubt a single Nema 17 could do that alone. I calculate it needs to lift 18.5 pounds of aluminum + weight of print........

You can still probably do it with a single large/long nema 17 if you use a low pitch lead screw (2mm pitch versus the 8mm I use) and pulley reduction. A 20 or even 16 tooth drive pulley and 36 or 40 tooth driven pulley on the 3 lead screws.

Here's a couple examples, though I know I've come across cleaner installs:
https://www.thingiverse.com/thing:1706163

https://www.thingiverse.com/thing:2185911

Parts for Adding Third Z Screw to D-Bot
Three lead screws and one nema 17 motor remix for C/D/J-bot
by oeyhaga

Yea I kinda knew it was ambitious, I was just like wow it would be so cool to be able to print stuff at 20" but as linear rods are completely out of the question and it will still be quite the stretch for liner guides I think I need to scale it down a bit.....

Do you think 16" or 18" for the X and Y is more realistic? Did you ever find out how the 16" printer went?

I think making it smaller is in my interest because having to limit print speeds is not really going to be in my benefit when printing at this size.

Thanks,
Taylor Dettling

Yes, I think 18" would be a good next step. You may be able to see some folks that have built larger printers on the facebook group: D-Bot / C-Bot Core-XY 3D Printer Group or by searching reddit. I know I've come across them before. There was also someone designing aluminum brackets to bolster the top brackets.
https://imgur.com/EWijlrW
https://imgur.com/UOMEues
If you are interested in these message "The Director" on Thingiverse
https://www.thingiverse.com/TheDirector/about

I also forgot to add that I am a little confused with the C-Bot cut length calculator, if I were to want a 16x16in build volume (or 400mm) in the X and Y axis and a 20 in build volume in the Z axis what values do I enter for the Z-Threaded rod, and bed dimensions? I am planning on using MIC-6 aluminum with a silicone heated bed attached to the bottom.

Good point, because leadscrews come in typical lengths it was setup that way. If you have a 300mm lead screw you will NOT have 300mm of build volume. I would guess there is 6-10cm of wastage depending on the hotend and carriage options you are using. If you want 20" of build height then I would put in a 515-520 leadscrew length.

Do you think it would be ok printed in PETG or should it be printed in ABS?

PETG will be great

Are you aware of anyone adapting and making a dual X carriage arrangement for either this or the D-bot?

Not that I am aware of. A dual parking extruder system would be interesting (with a drive carriage and magnetic coupling).

Any chance you could export/attach STEP files?

The C-bot was designed in 123D design which only allows you to export step files with the paid version. That said, another user has imported them and exported the step files with a better program. You can find them on the OpenBuilds site "files" section.

why aren't the bets on the sides fully concealed by the side bars? Is it a structural thing.

only half are enclosed as the other half must be exposed to engage with the idler pulleys on the XY bar

Why can't the pulleys be larger or pushed more towards the side bars

They could be larger with some redesign. Design uses 16 tooth pulleys and has the exposed belt space 1mm outside the extrusion.

I already started some redesigning https://tinkercad.com/things/afo1XIsFvtC
With the initial intention of greater biild volume to extrusion length ratio. But now I see the gain is negligible. So now it's more for esthetics. My main concern is the path of the belt must be clear of screws which would reduce stability. Do you think the loss of some screws will affect the stability of the machine?

I would try to maintain most of the screws in the upper frame as it is important to keep it rigid and square. If just one you are probably fine.

How large is it print size ? Cant find it anywhere

It is customizable. My original is 8"x12", but most people build 12"x12" version. 10"x10" has been done and someone tried 16"x16" (can't recall how that went). Height is also configurable and at least 36" has been done.
There is an extrusion length calculator in the BOM file of the OpenBuilds site. See the second tab.

Here are some pics of other builds: https://plus.google.com/u/0/photos/117092327090758898569/album/6312033886506017681?authkey=CLvxp5ie2uj_4wE&sqid=106707229610255461517&ssid=c20502c0-aaf3-4877-ac53-537dbbb43660

If your looking - the v-slot and some other parts are on sale right now.

*Edit -- Should have mentioned WHERE I found them - OpenBuildsPartStore.com. lol

do you have pics showing sample prints? preferably cubic objects

No really on hand. Lots of stuctural pieces printed with this printer and it handles 90 corners crisply if that is you concern. If you crank the speed up or put on the heavier direct drive extruder (all available on the openbuilds site) then you may need to clamp down on your acceleration limit.

what acceleration values do you use for direct drive? my concern was that how do the small wheels handle a direct drive extruder. Is it not two heavy for them?

I am still using bowden with accel at 6000 in slic3r (not sure in Marlin). Others are clamping down to 2-3000 with the heavier direct drive extruder. A new carriage that uses the E3D titan extruder is being created and is lighter with a pancake nema 17.
Small V wheels should be fine, you can use the extreme polycarbonate ones if you prefer.

Yes, check out the OpenBuilds site "drawing and files" tab. It has all the updated parts in a zip file. Should be in both the original and rework versions. http://www.openbuilds.com/builds/c-bot.1146/

I just ordered all of the parts online. Can't wait to start building!

Check out this excellent build blog by AK Eric: http://www.akeric.com/blog/?page_id=3063

Do you know where I can find the model files for the following parts shown? Thanks
http://i.imgur.com/Yc3aoIz.png

What are the approximate dimensions of the main v-slot rails? aka overall LxWxH

it is flexible depending on you bed size and lead screw length. Mine is about 14.5x22.5x18.5".

Thanks for the input, I didn't notice until today that there is a second tab of the BOM that explains all of the lengths, thanks for putting that on there

Well done! Do the printed corner assemblies provide enough rigidity to the frame? Also, do you you feel you're getting significantly better print quality with this coreXY design compared to your earlier, reprap-style printer?

Print quality wise it is better mainly due to using lead screws instead of threaded rod as I was getting some really good prints with my older OB1.4 printer aside from Z artifacts.
I can print faster and larger now, but neither of those mean better quality prints. The coreXY is very accurate in positioning but hotend selection and drive (bowden versus direct) are also important factors too.

The frame is fairly rigid for the 8x12 (200x300mm) bed size and 300mm lead screws. For larger beds and taller heights, guys have been boosting the rigidity at the lower half of the frame by using larger printed brackets or metal brackets from the open builds store. This helps the upper half too, but if you want to go really large then the upper plastic parts would need adjustment.

I really like the design overall. Have you had any issues at all with cantilevering the print bed like that? How heavy is the print bed? Thanks

If you really don't like the cantilevered bed, there is a mod on the Open Build site called the Triple C-bot that uses three lead screw intead of two so that it is not cantilevered.

No issues for me with my 8x12 bed. Some builders went with a larger (more cantilevered) bed and upgraded the 20x40 extrusion to 20x60 which you could do regardless if you were concerned. Didn't weigh it. It is a typical setup with PCB and glass on top.

Hi! Do you have video of working?

Check out the files tab at this link, http://www.openbuilds.com/builds/c-bot.1146/
Thingiverse would not let me post them here with a MOV extension.

Thanks for response! But http://www.openbuilds.com/builds/c-bot.1146/ not opening :(
Do you have youtube video? Youtube accept many formats for uploading.

Website must have been down, working now.

Hello thanks for sharing your awesome build with us. If I am reading the BOM right I need to print one of everything here with the exceptions of I need 10X 90_Corner.stl and 8x 3x3_Corner.stl right? Also it looks like the clamp parts on your hotend plate are hovering over the build plane. Not a issue if you print with support turned on but if you don't want to clean them you have to split the parts in the file. Again thanks for the awesome design I look forward to putting it together.

That sounds right for the part count. Make sure you have two XY bar ends.
Don't print those clamps with support, that would be a waste. If you are using slic3r just split the parts and they will all drop to the build plate. If not, the file version on the open build site has them lowered to the plate already.
Not sure if you've checked out the open build files, but there is a dual extruder carriage that works with E3Ds and J-heads that I would recommend. Also the modified Z axis clamps are nice, but not critical.

Comments deleted.

I really liked your design. It is very neat and I like the quality of the printed ends. What did you print those with?

This guy: thingiverse.com/make:61372
Though there have been a bunch of X-Axis modifications since the picture was taken

OB1.4 open beam 3D printer
by cfeniak

Cool design! Is there a discount for purchasing all the v-wheels needed? At $3.98 per wheel kit, that's roughly $80 for 20 wheels. That's the main obstacle preventing me from using V-slot, though I'd really like to try it. Unless you just buy the delrin part of the wheel and put your own bearings/screws in?

I'd say put a full order together with extrusion and such then ask them as there are no official discounts available. I got a cheaper shipping option by contacting them (I am in Canada and this was before they had Canadian distributors). I'd say a 5% savings might be achievable with a large order.
Parting it out is an option, but you still need the spacers and such so be careful. Also, I am not sure what bearings the mini Vs use.

The bearings look to be 608 series. It's the most commonly used so far for 3d printers.

608 bearings have a 8mm ID. All the V-slot wheels run on 5mm.

My apologies it's a 625 bearing I mean. Sorry I have both 608 and 625 in front of me.

It's a very tidy build. You've inspired me to do similar with 20x40 vslot I have for a printer project. I like your corner bracket/ pulley support/ endstop support combo - clever. How long is the belt? Do you have any more detail explaining the belt routing? I can't quite get my head around it. Have you got anything on CAD?
Thanks for sharing.
Alan

Have you checked out the open builds site? There are some extra photos in the discussion too.
It is a coreXY setup just like: http://www.corexy.com/theory.html
However, the belts "cross over" in the reference mechanism I linked to at the top/back. That ends up looking like this: http://www.thingiverse.com/thing:22005 The cross over is required here as the belts run in the same XY plane.
I wanted to avoid the cross over so I offset the two belt runs by 20mm. This also enabled me run the belts partially through the extrusion. You can see this by looking at the hotend carriage and looking where the belts attach on the left versus right side.

CoreXY
by frankie

Thanks for the answer. I did figure it out eventually!
Nice work Carl.

-Alan

I was wanting to do this as I really hate my I3 rework in certain areas. No 8mm smooth rods, etc... would be awesome and should make it damn strong, and sturdy too.

If my budget ever allows it I am going to make a V-Rail 3d printer as well because I think the days of smooth rods are dead. Heck, I got my Chrome hardened 8mm smooth rods (due to how wet it is here I have to use anything but plain steel) from Mitsumi and they were free at the start of 2014 due to a 150 dollar promotion but had they not been I would not have a printer right now. With V-Rail the prices are less from what I added up and easier to build. The hardest thing is making sure the cuts are EXACTLY 90 degrees and smooth as V-Rail does not come pre cut unlike my smooth rods.

I like that the V-Slot is dual purpose and can be used both structurally and as a guide for linear motion. Reduces cost and can simplify a build.
A perfect 90 cut would be great... but isn't going to happen in typical home setups. Fortunately the brackets can accommodate these imperfections as the exterior clamping on the uncut surfaces defines their alignment and not the cut end surface. Small differences in length can be adjusted for by slightly moving the extrusion in and out of the bracket to easily square off the frame. That is an advantage of the V slot compared to a defined hole location for positioning.

Well, this is why I never made my CNC router because I don't have the tools to cut perfectly and in the OX build, or maybe it was Routy, they said it has to be cut perfectly. I don't own a chop saw and even if I did I wouldn't have an 80 tooth 50 dollar blade for it.

I was thinking of one improvement for the I3 is to remove those X bars and replace with V-Rail as the I3 has a bad issue of the hot end moving due to the way it was made (it can move up to 1mm which is a HUGE run out when trying to dial everything in). It would take some re-doing of the plastic pieces but noting I couldn't manage. Three plastic pieces would need to be redone for that but the next issue would mean redoing it all from scratch and that is the horrible Y and I could never get it level after over 200, yes two hundred, hours of trying. I gave up and went with ABL for that but with V-Rail I think that issue would be non existent.